Virtual silicon generation system and generation method

Abstract

The application provides a virtual silicon chip generation system and a generation method. The virtual silicon chip generation system comprises a data collection unit and a silicon chip generation unit. The data collection unit is used for collecting feature information and process information of a silicon chip from a detection database. The silicon chip generation unit is used for generating a virtual silicon chip according to the feature information and the process information. Thus, the application not only utilizes the feature information of the silicon chip in the detection database, but also utilizes the process information in the preparation process of the silicon chip, so that the generated virtual silicon chip is closer to the real silicon chip, can fully reflect the information on the real silicon chip and the real situation of the process, and improves the use value of the virtual silicon chip.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of integrated circuit manufacturing, in particular to a virtual wafer generation system and a generation method. BACKGROUND

[0002] The virtual wafer is a digital wafer corresponding to the real wafer to be detected, which is generated based on process data and result data of the detection equipment. The virtual wafer can present the process condition of the real wafer and the optical imaging process of the detection equipment in the form of a digital image. In actual production process, it has a wide range of application scenarios, for example: the virtual wafer can reproduce the on-site problems of the Fab factory, and simulate the detection of the virtual wafer according to the optical parameters and algorithm parameters used on site. The on-site problems can be located remotely in an offline state, reducing the on-site maintenance cost. Moreover, since the virtual wafer is generated based on the existing optical imaging and algorithm process, the existing algorithm can be modified or improved based on the virtual wafer, and the difference in function and performance between the new algorithm and the existing algorithm is compared to realize remote development of the new algorithm in an offline state, reducing the algorithm development cost. In addition, due to the special confidentiality requirements of the Fab factory, the virtual wafer has become an essential tool for the development and real-time processing of on-site problems of the detection equipment algorithm.

[0003] However, the existing virtual wafer generation system is based on the analysis of physical data rather than the analysis of virtual models, that is, there is often a lack of connection between the physical space and the virtual space of the product, so that the virtual wafer lacks the process characteristic information of the real wafer, resulting in a lack of authenticity of the virtual wafer. This will naturally bring disadvantages to the subsequent research and development and problem troubleshooting.

[0004] Therefore, a new virtual wafer generation system is needed to enhance the authenticity of the virtual wafer. SUMMARY

[0005] The present application aims to provide a virtual wafer generation system and a generation method to solve the problem of how to improve the authenticity of the virtual wafer.

[0006] To solve the above technical problems, the present application provides a virtual wafer generation system, comprising: a data collection unit and a wafer generation unit; wherein,

[0007] The data collection unit is used to collect feature information and process information of the wafer from a detection database;

[0008] The wafer generation unit is used to generate a virtual wafer according to the feature information and the process information.

[0009] Optionally, in the virtual wafer generating system, the virtual wafer generating system further comprises an image information unit; the image information unit comprises a standard map data set, a difference map data set and a defect map data set; the standard map data set is used to represent standard wafer information; the difference map data set is used to represent wafer information within a maximum allowable range of process parameter variation; the defect map data set is used to represent defect information on the wafer.

[0010] The standard map data set is generated according to the feature information; the difference map data set and the defect map data set are generated according to the feature information and / or the process information; and the wafer generating unit is used to generate a virtual wafer according to the standard map data set, the difference map data set and the defect map data set.

[0011] Optionally, in the virtual wafer generating system, the feature information comprises periodic structure information on the wafer; and the process information comprises process parameters, process parameter thresholds, process procedure generated data and location information of the periodic structure on the wafer.

[0012] Optionally, in the virtual wafer generating system, the process procedure generated data comprises feature values and distribution data of key areas, difference information between at least two grain images, attributes of defects, location information of defects and classification rules of defects.

[0013] Optionally, in the virtual wafer generating system, the periodic structure information comprises pictures of the periodic structure and / or pictures of feature areas in the periodic structure, or pictures obtained by statistics of multiple pictures; the pictures obtained by statistics comprise mean value maps or median value maps.

[0014] Optionally, in the virtual wafer generating system, the standard map data set comprises pictures of the periodic structure without defects or a combined picture of multiple pictures.

[0015] Optionally, in the virtual wafer generating system, the feature information and the process information both comprise chip design data.

[0016] Optionally, in the virtual wafer generating system, the standard map data set comprises standard pictures generated by simulation of the chip design data, or a combined picture of multiple standard pictures.

[0017] Optionally, in the virtual wafer generating system, the difference map data set comprises pictures generated by simulation of the periodic structure information and the process parameter thresholds, and the pictures generated by simulation are arranged according to the location information of the periodic structure on the wafer.

[0018] Optionally, in the virtual wafer generation system, the difference map dataset comprises a picture of the periodic structure after operation processing; and the operation processing method comprises standard deviation operation and variance operation.

[0019] Optionally, in the virtual wafer generation system, the defect map dataset comprises a picture generated by the periodic structure information and data simulation generated by the process.

[0020] Optionally, in the virtual wafer generation system, the gray value of any pixel point in the picture in the defect map dataset satisfies the condition that the gray value is greater than the sum of the gray value of the pixel point at the corresponding position in the picture in the standard map dataset and the gray value of the pixel point at the corresponding position in the picture in the difference map dataset, or the gray value is less than the difference between the gray value of the pixel point at the corresponding position in the picture in the standard map dataset and the gray value of the pixel point at the corresponding position in the picture in the difference map dataset.

[0021] Optionally, in the virtual wafer generation system, the wafer generation unit comprises a first random number generator and a second random number generator; and the gray value of any pixel point on the virtual wafer satisfies the following formula:

[0022] When the pixel point is defect-free, the following formula is satisfied:

[0023] Gray_Digital_wafer(x,y)=Gray_Standard(x,y)+R1*Gray_Std(x,y);

[0024] When the pixel point has a defect, the following formula is satisfied:

[0025] Gray_Digital_wafer(x,y)=Gray_Standard(x,y)+R2*Gray_Defect(x,y);

[0026] wherein,

[0027] Gray_Digital_wafer(x,y) is the gray value of the pixel point (x, y) on the virtual wafer;

[0028] Gray_Standard(x,y) is the gray value of the pixel point (x, y) in the picture in the standard map dataset;

[0029] Gray_Std(x,y) is the gray value of the pixel point (x, y) in the picture in the difference map dataset;

[0030] Gray_Defect(x,y) is the gray value of the picture in the defect map data set at pixel position (x, y);

[0031] R1 is a first random number generator; R2 is a second random number generator.

[0032] Optionally, in the virtual silicon wafer generation system, the data collection unit also collects hardware device information; the hardware device information includes detection indexes and hardware attributes of the detection device imaging system.

[0033] Optionally, in the virtual silicon wafer generation system, the virtual silicon wafer generation system further comprises a silicon wafer judging unit for judging whether the generated virtual silicon wafer meets the set requirements.

[0034] Optionally, in the virtual silicon wafer generation system, the silicon wafer judging unit comprises a silicon wafer data set and a judging statistical data set; the silicon wafer data set contains characteristic information of a silicon wafer compared with the virtual silicon wafer; and the judging statistical data set contains process information and comparison results of the comparison of the virtual silicon wafer with the silicon wafer.

[0035] Optionally, in the virtual silicon wafer generation system, the characteristic information includes pictures of periodic structures in the silicon wafer and arrangement orders of the pictures of periodic structures.

[0036] The process information includes comparison difference value maps obtained by subtracting gray values of the pictures of periodic structures on the virtual silicon wafer from gray values of corresponding pictures of periodic structures in the silicon wafer according to the arrangement orders, and statistical data of all the comparison difference value maps.

[0037] The comparison results include judging results of whether each comparison difference value map meets a preset normal distribution.

[0038] Based on the same inventive concept, the present application further provides a virtual silicon wafer generation method, comprising:

[0039] A detection device detects a plurality of silicon wafer information and stores the silicon wafer information in a detection database;

[0040] A data collection unit collects characteristic information and process information of silicon wafers from the detection database;

[0041] A silicon wafer generation unit generates a virtual silicon wafer according to the characteristic information and the process information.

[0042] Optionally, in the virtual silicon wafer generation method, the data collection unit also collects hardware device information in the detection database.

[0043] Optionally, in the method for generating the virtual wafer, after the virtual wafer is generated, the method for generating the virtual wafer further comprises:

[0044] The wafer judging unit is used to judge whether the virtual wafer meets the set requirements, if yes, the generation of the virtual wafer is completed; if no, the related process is modified according to the nature of the abnormal point, and then the virtual wafer is generated again and judged until the set requirements are met.

[0045] In summary, the present application provides a virtual wafer generation system and a generation method. The virtual wafer generation system comprises a data collection unit and a wafer generation unit. The data collection unit is used to collect feature information and process information of a wafer from a detection database. The wafer generation unit is used to generate a virtual wafer according to the feature information and the process information. Thus, the present application not only utilizes the feature information of the wafer in the detection database, but also utilizes the process information in the preparation process of the wafer, so that the generated virtual wafer is closer to the real wafer, can fully reflect the information on the real wafer and the real situation of the process, and improves the use value of the virtual wafer. BRIEF DESCRIPTION OF DRAWINGS

[0046] Figure 1 is a structure schematic diagram of a virtual wafer generation system in an embodiment of the present application;

[0047] Figure 2 is a structure schematic diagram of a periodic structure in an embodiment of the present application;

[0048] Figure 3 is a structure schematic diagram of a detection imaging system in an embodiment of the present application;

[0049] Figure 4 is a structure schematic diagram of a virtual wafer generation system in an embodiment of the present application;

[0050] Figure 5 is a schematic diagram of a combination image with the largest gray value in an embodiment of the present application;

[0051] Figure 6 is a schematic diagram of a combination image with the smallest gray value in an embodiment of the present application;

[0052] Figure 7 is a structure schematic diagram of a virtual wafer in an embodiment of the present application;

[0053] Figure 8 is a schematic diagram of a virtual wafer generation process in an embodiment of the present application;

[0054] Figure 9 is a flowchart of a virtual wafer generation method in an embodiment of the present application;

[0055] Wherein, the reference signs are:

[0056] 10 - data collection unit; 20 - image information unit; 201 - standard map data set; 202 - difference map data set; 203 - defect map data set 203; 30 - wafer generation unit; 40 - wafer determination unit;

[0057] P - periodic structure; D - die D; S - scribe lane; M - feature area; N - virtual wafer;

[0058] C101 - bright field camera; C102 - dark field camera; C103 - bright field detection camera; C104 - bright field illumination; C105 - coaxial autofocus; C106 - objective lens; C107 - dark field illumination. DETAILED DESCRIPTION

[0059] In order to make the objectives, advantages and features of the present application clearer, the following further describes the present application in combination with the drawings and specific embodiments. It should be noted that the drawings are very simplified and not drawn according to scale, and are only used to facilitate and clarify the purpose of assisting the description of the embodiments of the present application. In addition, the structures shown in the drawings are often part of the actual structures. In particular, the emphasis of each drawing needs to be different, and sometimes different scales are used. It should also be understood that, unless specifically described or indicated, the terms "first", "second", "third" and the like in the description are only used to distinguish the components, elements, steps and the like in the description, and are not used to represent the logical relationship or sequence relationship between the components, elements, steps and the like.

[0060] To solve the above technical problems, the present embodiment provides a virtual wafer generation system. Please refer to Figure 1 , the virtual wafer generation system comprises a data collection unit 10 and a wafer generation unit 30; wherein the data collection unit 10 is used to collect feature information and process information of a wafer from a detection database; the wafer generation unit 30 is used to generate a virtual wafer according to the feature information and the process information. Therefore, adding the process information in the real wafer preparation process in the generation process of the virtual wafer can make the generated virtual wafer closer to the real wafer, can fully reflect the information on the real wafer and the real situation of the process, and improves the use value of the virtual wafer.

[0061] Furthermore, the virtual silicon wafer generation system also includes an image information unit 20. The image information unit 20 includes a standard image dataset 201, a difference image dataset 202, and a defect image dataset 203. The standard image dataset 201 is used to characterize standard silicon wafer information, i.e., silicon wafer information in its theoretical state, without defects in the process; the difference image dataset 202 is used to characterize silicon wafer information within the maximum allowable range of process parameter variations, i.e., information about silicon wafers prepared within the critical range of various process parameter thresholds; the defect image dataset 203 is used to characterize defect information on the silicon wafer; wherein, the standard image dataset 201 is generated based on the feature information; the difference image dataset 202 and the defect image dataset 203 are generated based on the feature information and / or the process information; the silicon wafer generation unit 30 is used to generate a virtual silicon wafer based on the standard image dataset 201, the difference image dataset 202, and the defect image dataset 203.

[0062] The testing database is generated by the testing equipment during the silicon wafer testing process. The silicon wafers can be process wafers that have completed all processes, or test wafers can be used alone to complete all processes to qualify as silicon wafers. This embodiment does not limit the number of silicon wafers; it can be 100, 150, or 200 wafers, etc.

[0063] Furthermore, the data collection unit 10 is used to collect various information about the silicon wafer, including a periodic information dataset 101 and a process information dataset 102. The characteristic information of the silicon wafer belongs to the periodic information dataset 101, and the process information belongs to the process information dataset 102. The characteristic information of the silicon wafer includes periodic structure information on the silicon wafer. This can be an image of the periodic structure, and / or an image of a characteristic region within the periodic structure, or a statistically analyzed image of multiple images, such as a mean plot or a median plot. Specifically, for example... Figure 2 As shown, the pitch P can be selected as the periodic structure. The pitch P is the smallest periodic structure on the silicon wafer, including the die D and the dicing groove S. Furthermore, a specific portion M within the die D can be selected as the feature region to characterize the silicon wafer's feature information. In addition, to fully characterize the silicon wafer's feature information, the periodic information dataset 101 can also include chip design data, such as chip design diagrams. Therefore, the data storage format of the data collection unit 10 can be text format or digital image format, such as CAD or GDS images.

[0064] The process information is used to reflect the relevant situation of the silicon wafer during the manufacturing process. This process information includes process parameters, process parameter thresholds, data generated during the process, and the position information of the periodic structure on the silicon wafer. The data generated during the process includes, but is not limited to: feature values ​​and distribution data of the care area, difference information between at least two grain images, defect attributes, defect location information, and defect classification rules. Similarly, the process information may also include chip design data. Therefore, this embodiment improves the realism of the subsequently generated virtual silicon wafer N by adding a large amount of process information data, which is beneficial for evaluating pre-research products, verifying and developing various algorithms, and identifying and detecting various R&D problems.

[0065] Furthermore, to further enhance the realism of the virtual silicon wafer N, the data collection unit 10 also collects hardware device information; this hardware device information includes the detection indicators and hardware attributes of the detection device imaging system. The detection device imaging system includes an optical imaging device and a photoelectric conversion device. Further, as... Figure 3 The imaging system of the detection equipment shown includes various lenses, a bright-field camera C101, a dark-field camera C102, a bright-field detection camera C103, a bright-field illumination C104, a coaxial autofocus C105, an objective lens C106, and a dark-field illumination C107. Therefore, the hardware equipment information may include bright-field and dark-field illumination information, lens magnification, etc. For this reason, as... Figure 4 As shown, the data collection unit 10 further includes a hardware information dataset 103; the hardware device information belongs to the hardware information dataset 103.

[0066] For further details, please refer to Figure 1 and 4 The image information unit 20 is used to generate the feature information required for the virtual silicon wafer N based on the data provided by the data collection unit 10. The standard image dataset 201 in the image information unit 20 serves as the information source for the theoretical silicon wafer, containing images of the periodic structure with defects removed, or combinations of multiple images. Figure 5The shown combination picture is spliced by 4 pictures of periodic structure. Since the standard picture data set 201 is the design standard of the chip, when the standard picture in the standard picture data set 201 is simulated by the chip design data in the periodic information data set 101, or the combination picture of multiple standard pictures, the picture in the standard picture data set 201 is free of defects and meets the set requirements. When the picture of the periodic structure of the silicon wafer in the periodic information data set 101 is used, some defects inevitably exist due to the self or process, and thus image processing needs to be performed on the corresponding picture or combination picture to remove the defective abnormal points, so as to ensure that the picture or combination picture in the standard picture data set 201 is free of defects and is an idealized data picture.

[0067] The difference picture data set 202 is used to reflect the maximum tolerance range of each process parameter change in the chip manufacturing process. In other words, the difference picture data set 202 contains information within the threshold critical point of each process parameter. Further, the difference picture data set 202 is one or more digital pictures calculated based on the periodic structure picture or the combination picture of multiple periodic structures and process-related information, which can also include hardware device information. The difference picture data set 202 can also be the threshold distribution of detection algorithms of different regions generated according to the texture density and shape of the chip design picture in the CAD or GDS format. The difference picture data set 202 contains the picture simulated by the periodic structure information and the process parameter threshold. Further, the picture in the difference picture data set 202 can be simulated by the picture or combination picture in the standard picture data set 201 and the process parameter threshold, and the simulated picture is arranged according to the position information of the periodic structure on the silicon wafer.

[0068] In addition, the difference picture data set 202 also contains the picture of the periodic structure after operation processing; wherein the operation processing method includes standard deviation operation and variance operation. Further, the picture in the difference picture data set 202 can be obtained by performing standard deviation or variance operation processing on the picture or combination picture in the periodic information data set 101, or performing standard deviation or variance operation processing with a set multiple. Specifically, the picture in the difference picture data set 202 can be a standard deviation picture of digital images of multiple dies D on the same silicon wafer, or a standard deviation picture of digital images of multiple dies D on the same silicon wafer with any multiple. Or, as shown in the figure, the picture in the difference picture data set 202 is a standard deviation picture of digital images of multiple dies D on the same silicon wafer with a set multiple. Figures 5-6 As shown in the figure, since the picture in the difference picture data set 202 is a picture within the threshold range of the parameter, the maximum value and the minimum value of the gray value are compared, Figure 5 the maximum value of the gray value is Figure 6 the minimum value of the gray value. Therefore, Figure 5 , Figure 6 and in Figure 5 and6 The gray scale values between the pictures correspond to the difference map data set 202. Therefore, the difference map data set 202 can be used as a source of information for qualified silicon wafers within the maximum allowable range of process parameters.

[0069] The defect map data set 203 is one or more pixel point sets corresponding to the pixel point positions of the pictures in the standard map data set 201. The defect map data set 203 mainly represents the defect information on the silicon wafer. Of course, in the ideal case without defects, the defect map data set 203 can also be an empty set. Further, the defect map data set 203 contains the pictures generated by the data simulation of the periodic structure information and the process generated. Specifically, the pictures in the standard map data set 201 or the combined map and the pictures generated by the simulation of the attributes of the defects, the position information of the defects, and the classification rules of the defects are the pictures in the defect map data set 203. Moreover, the gray scale value of any pixel point in the pictures in the defect map data set 203 satisfies the condition that it is greater than the sum of the gray scale value of the pixel point at the corresponding position in the picture in the standard map data set 201 and the gray scale value of the pixel point at the corresponding position in the picture in the difference map data set 202, or it is less than the difference between the gray scale value of the pixel point at the corresponding position in the picture in the standard map data set 201 and the gray scale value of the pixel point at the corresponding position in the picture in the difference map data set 202. In other words, the gray scale value of the picture in the defect map data set 203 is outside the threshold range, and the gray scale value of each pixel point is an arbitrary value that meets the classification rules.

[0070] Further, referring to Figure 7 , the silicon wafer generating unit 30 is configured to generate the virtual silicon wafer N according to the data provided by the image information unit 20. The silicon wafer generating unit 30 includes a first random number generator and a second random number generator. Moreover, the gray scale value of any pixel point on the virtual silicon wafer N satisfies the following formula:

[0071] When the pixel point has no defects, the following formula is satisfied:

[0072] Gray_Digital_wafer(x,y) = Gray_Standard(x,y) + R1*Gray_Std(x,y);

[0073] When the pixel point has defects, the following formula is satisfied:

[0074] Gray_Digital_wafer(x,y) = Gray_Standard(x,y) + R2*Gray_Defect(x,y);

[0075] wherein,

[0076] Gray_Digital_wafer(x,y) is the gray value of the pixel position (x, y) on the virtual silicon wafer N;

[0077] Gray_Standard(x,y) is the gray value of the pixel position (x, y) on the picture in the standard map data set 201;

[0078] Gray_Std(x,y) is the gray value of the pixel position (x, y) on the picture in the difference map data set 202;

[0079] Gray_Defect(x,y) is the gray value of the pixel position (x, y) on the picture in the defect map data set 203;

[0080] R1 is a first random number generator; R2 is a second random number generator.

[0081] It can be understood that the silicon wafer generating unit 30 takes the picture in the standard map data set 201 as a reference. For the part without defects, data is randomly obtained from the difference map data set 202 for supplementation; and for the part with defects, data is randomly obtained from the defect map data set 203 for supplementation, so as to generate the complete virtual silicon wafer N. Since the virtual silicon wafer N covers the characteristic information, process information and hardware device information of the silicon wafer processed by the real process, the virtual silicon wafer N has better authenticity, greatly improves the use value of the virtual silicon wafer N, and further improves the research and development efficiency and reduces the cost.

[0082] Of course, the selection mode of the random number has a certain uncertainty, which may cause the authenticity of the virtual silicon wafer N to deviate. Therefore, in order to further improve the authenticity of the virtual silicon wafer N, the virtual silicon wafer generating system provided in the embodiment further comprises a silicon wafer judging unit 40. Please refer to Figure 8 The judging unit 40 is used for judging whether the generated virtual silicon wafer N meets the set requirements, so as to modify according to the judgment result, so as to further improve the authenticity of the virtual silicon wafer N.

[0083] Further, the wafer judging unit 40 comprises a wafer data set and a judging statistic data set. The wafer data set comprises feature information of the wafer compared with the virtual wafer N; the judging statistic data set comprises process information and comparison results of the virtual wafer N compared with the wafer. The feature information of the wafer comprises pictures of all periodic structures in the wafer and arrangement order of the pictures. In other words, the selected wafer is a target real wafer to be simulated by the virtual wafer N. The process information comprises comparison difference value maps obtained by subtracting gray values of the pictures of periodic structures on the virtual wafer N from corresponding gray values of the pictures of periodic structures in the wafer according to the arrangement order of the pictures, and data obtained by statistical processing of all the comparison difference value maps. The statistical processing methods include, but are not limited to, variance, standard deviation or mean value, etc.

[0084] Further, the comparison results comprise judging results of whether each comparison difference value map satisfies a preset normal distribution of mean value and variance, so that according to the judging results, process-related information is supplemented or modified according to the properties of abnormal points, the virtual wafer N is regenerated, and the judging is performed again until the virtual wafer N meets the set requirements. Therefore, the wafer judging unit added in the embodiment can judge the authenticity of the generated virtual wafer N. If abnormal points deviating from the authenticity greatly appear on the virtual wafer N due to process parameters, etc., the properties of the abnormal points can be extracted in time, and the reasons are analyzed, and then the generation process of the virtual wafer N is modified by gradually adjusting the process parameters, so that the virtual wafer N meeting the judging requirements of the wafer judging unit 40 can be finally obtained. It can be seen that the wafer judging unit 40 is an automatic check of the authenticity degree of the generated virtual wafer N, and can further improve the authenticity of the finally obtained virtual wafer N.

[0085] Based on the same inventive concept, the embodiment further provides a virtual wafer generation method. As shown in Figures 8-9 The virtual wafer generation method comprises:

[0086] Step S10: a detection device detects a plurality of wafer information and stores the wafer information in a detection database.

[0087] Further, the wafer information includes feature information of the wafer, process information, and hardware device information. The feature information includes periodic structure information on the wafer. The periodic structure information can be a picture of the periodic structure, and / or a picture of a feature region in the periodic structure, or a picture obtained by statistical processing of multiple pictures, such as a mean value picture or a median value picture. The process information includes process parameters, process parameter threshold values, data generated in a process, and position information of the periodic structure on the wafer. The hardware device information includes detection indexes and hardware attributes of an imaging system of a detection device.

[0088] Step two S20: The data collection unit 10 collects feature information and process information of the wafer from the detection database. Meanwhile, the data collection unit 10 also collects hardware device information.

[0089] The data collection unit 10 includes a periodic information data set 101, a process information data set 102, and a hardware information data set 103. The feature information of the wafer belongs to the periodic information data set 101, the process information belongs to the process information data set 102, and the hardware device information belongs to the hardware information data set 103. The feature information is used by the graphic information unit 20 to generate a standard picture data set 201, and is used by the graphic information unit 20 to generate a difference picture data set 202 and a defect picture data set 203 respectively based on the standard picture data set 201 or the feature information and the process information.

[0090] Further, the pictures in the standard picture data set 201 are processed to remove defect abnormal points, and are used to represent standard wafer information. The difference picture data set 202 includes pictures obtained by simulating the pictures in the standard picture data set 201 or the combined picture with the process parameter threshold values, and the simulated pictures are arranged according to position information of the periodic structure on the wafer. The difference picture data set 202 is used to represent wafer information within a maximum allowable range of process parameter changes. The defect picture data set 203 includes pictures obtained by simulating the pictures in the standard picture data set 201 or the combined picture with defect attributes, position information of the defects, and classification rules of the defects. The defect picture data set 203 is used to represent defect information on the wafer.

[0091] Step three S30: A wafer generation unit 30 generates a virtual wafer N based on the feature information and the process information. That is, the wafer generation unit 30 generates the virtual wafer N based on the standard picture data set 201, the difference picture data set 202, and the defect picture data set 203, so that the generated virtual wafer N includes information of the wafer in various states, and is closer to a real wafer, thereby being able to fully reflect real conditions of devices on the wafer and a process, and improving the use value of the virtual wafer N.

[0092] wherein the gray value of any pixel point on the virtual wafer N formed satisfies the following formula:

[0093] When the pixel point has no defect, it satisfies:

[0094] Gray_Digital_wafer(x,y)=Gray_Standard(x,y)+R1*Gray_Std(x,y);

[0095] When the pixel point has a defect, it satisfies:

[0096] Gray_Digital_wafer(x,y)=Gray_Standard(x,y)+R2*Gray_Defect(x,y);

[0097] wherein,

[0098] Gray_Digital_wafer(x,y) is the gray value of the pixel position (x, y) on the virtual wafer N;

[0099] Gray_Standard(x,y) is the gray value of the pixel position (x, y) on the picture in the standard map data set 201;

[0100] Gray_Std(x,y) is the gray value of the pixel position (x, y) on the picture in the difference map data set 202;

[0101] Gray_Defect(x,y) is the gray value of the pixel position (x, y) on the picture in the defect map data set 203;

[0102] R1 is a first random number generator; R2 is a second random number generator.

[0103] After the virtual wafer N is generated, the virtual wafer generation method further comprises:

[0104] using a wafer judgment unit 40 to judge whether the virtual wafer N meets the set requirements, if yes, the generation of the virtual wafer N is completed; if no, the related process is modified according to the nature of the abnormal point, and the virtual wafer N is generated again for judgment until the set requirements are met.

[0105] Specifically, a silicon wafer is selected as a reference object, wherein the silicon wafer is a real silicon wafer to be simulated by the virtual silicon wafer N. Then, the gray values of the pictures of the periodic structures on the virtual silicon wafer N and the corresponding pictures of the periodic structures in the silicon wafer are subtracted in the same arrangement order, so as to obtain a comparison difference graph, and a comparison difference graph sequence is generated. Secondly, it is verified whether each comparison difference graph meets the normal distribution of the set mean and variance. If the comparison difference graph data meets the normal distribution of the set mean and variance, it can be considered that the virtual silicon wafer N is within an acceptable range; if the comparison difference graph data does not meet the normal distribution of the set mean and variance, process-related information needs to be supplemented or modified according to the nature of the abnormal points, and the virtual silicon wafer N is regenerated until the comparison difference graph data meets the normal distribution of the set mean and variance, so as to obtain a virtual silicon wafer with better authenticity.

[0106] In summary, the virtual silicon wafer generation system and the generation method provided by the embodiment are provided. The virtual silicon wafer generation system includes the data collection unit 10, the image information unit 20, and the silicon wafer generation unit 30. The image information unit 20 includes a standard graph data set 201, a difference graph data set 202, and a defect graph data set 203. The standard graph data set 201 is used to represent standard silicon wafer information. The difference graph data set 202 is used to represent silicon wafer information within the maximum allowable range of process parameters. The defect graph data set 203 is used to represent defect information on the silicon wafer. The standard graph data set 201 is generated according to the characteristic information. The difference graph data set 202 and the defect graph data set 203 are generated according to the characteristic information and / or the process information. The silicon wafer generation unit 30 is used to generate a virtual silicon wafer N according to the standard graph data set 201, the difference graph data set 202, and the defect graph data set 203. In other words, the standard graph data set 201 is the information source of the theoretical silicon wafer, the difference graph data set 202 is the information source of the qualified silicon wafer within the maximum allowable range of process parameters, and the defect graph data set 203 is the information source of the defect situation of the silicon wafer. The virtual silicon wafer N obtained according to the standard graph data set 201, the difference graph data set 202, and the defect graph data set 203 can include information of the silicon wafer in various states, so that the generated virtual silicon wafer N is closer to the real silicon wafer, and can fully reflect the real situation of the device and the process on the silicon wafer, thereby improving the use value of the virtual silicon wafer N.

[0107] Furthermore, it should be appreciated that the above-described embodiments are merely exemplary as to the present application and need not be practiced as described. It should also be understood that where the above-described embodiments are, for clarity, described in terms of steps or means, these steps or means need not be performed in the order given or even one at a time. Rather, certain steps or means can be performed in an order different than other steps or means or one following another, or performed at least at partially in parallel, to express the sequential order of individual steps or features or functions. Also, description of the above-described embodiments is meant to be illustrative only and not limiting as to the scope of the present application, which is set forth in the following claims.

Claims

1. A virtual silicon generation system, comprising: The system comprises: a data collection unit and a wafer generation unit; wherein the data collection unit is configured to collect feature information and process information of a wafer from a detection database; the wafer generation unit is configured to generate a virtual wafer according to the feature information and the process information; the feature information comprises periodic structure information on the wafer; and the process information comprises process parameter threshold values; and the virtual wafer generation system further comprises an image information unit; the image information unit comprises a difference map data set; the difference map data set comprises pictures generated by simulating the periodic structure information and the process parameter threshold values, and the pictures generated by simulation are arranged according to position information of the periodic structure on the wafer.

2. The virtual silicon generation system of claim 1, wherein, The image information unit further comprises a standard map data set and a defect map data set; the standard map data set is configured to represent standard wafer information; the difference map data set is configured to represent wafer information within a maximum allowable range of process parameter variation; and the defect map data set is configured to represent defect information on the wafer. The standard map data set is generated according to the feature information; and the difference map data set and the defect map data set are generated according to the feature information and / or the process information. The wafer generation unit is configured to generate a virtual wafer according to the standard map data set, the difference map data set and the defect map data set.

3. The virtual silicon generation system of claim 2, wherein, The process information further comprises process parameters, process procedure generated data and position information of the periodic structure on the wafer.

4. The virtual silicon generation system of claim 3, wherein, The process procedure generated data comprises feature values and distribution data of critical areas, difference information between at least two grain images, attributes of defects, position information of defects and classification rules of defects.

5. The virtual silicon generation system of claim 3, wherein, The periodic structure information comprises pictures of the periodic structure and / or pictures of feature areas in the periodic structure, or pictures obtained by statistics of multiple pictures; wherein the pictures obtained by statistics comprise mean value pictures or median value pictures.

6. The virtual silicon generation system of claim 3, wherein, The standard map data set comprises pictures of the periodic structure without defects or combined pictures of multiple pictures.

7. The virtual silicon generation system of claim 3, wherein, Chip design data is included in the feature information and the process information.

8. The virtual silicon generation system of claim 7, wherein, The standard map data set comprises standard pictures generated by simulation of the chip design data, or combined pictures of multiple standard pictures.

9. The virtual silicon generation system of claim 1, wherein, The difference map data set comprises pictures of the periodic structure after operation processing; wherein the operation processing method comprises standard deviation operation and variance operation.

10. The virtual silicon generation system of claim 3, wherein, The defect map data set comprises pictures generated by simulation of the periodic structure information and the process procedure generated data.

11. The virtual silicon generation system of claim 10, wherein, The gray value of any pixel in the pictures in the defect map data set satisfies the condition that the gray value is greater than the sum of the gray value of the pixel at the corresponding position in the pictures in the standard map data set and the gray value of the pixel at the corresponding position in the pictures in the difference map data set, or the gray value is less than the difference between the gray value of the pixel at the corresponding position in the pictures in the standard map data set and the gray value of the pixel at the corresponding position in the pictures in the difference map data set.

12. The virtual silicon generation system of claim 2, wherein, The silicon wafer generating unit comprises a first random number generator and a second random number generator; and a gray value of any pixel point on the virtual silicon wafer satisfies the following formula: When the pixel point is defect-free, it satisfies: Gray_Digital_wafer(x, y)= Gray_ Standard (x, y)+R1*Gray_ Std(x, y); When the pixel point has a defect, it satisfies: Gray_Digital_wafer(x, y)= Gray_ Standard(x, y)+R2*Gray_ Defect(x, y); wherein, Gray_Digital_wafer(x, y) is a gray value of a pixel position (x, y) on the virtual silicon wafer; Gray_ Standard (x, y) is a gray value of a pixel position (x, y) on a picture in the standard image dataset; Gray_ Std (x, y) is a gray value of a pixel position (x, y) on a picture in the difference image dataset; Gray_ Defect(x,y) is a gray value of a pixel position (x, y) on a picture in the defect image dataset; R1 is a first random number generator; and R2 is a second random number generator.

13. The virtual silicon generation system of claim 1, wherein, The data collection unit also collects hardware device information; the hardware device information comprises detection indexes and hardware attributes of a detection device imaging system.

14. The virtual silicon generation system of claim 1, wherein, The virtual silicon wafer generating system further comprises a silicon wafer judging unit for judging whether the generated virtual silicon wafer satisfies a set requirement.

15. The virtual silicon generation system of claim 14, wherein, The silicon wafer judging unit comprises a silicon wafer dataset and a judging statistical dataset; the silicon wafer dataset contains characteristic information of a silicon wafer compared with the virtual silicon wafer; and the judging statistical dataset contains process information and comparison results of the virtual silicon wafer compared with the silicon wafer.

16. The virtual silicon generation system of claim 15, wherein, The characteristic information comprises pictures of periodic structures in the silicon wafer and arrangement orders of the pictures of periodic structures; The process information comprises comparison difference value images obtained by subtracting gray values of the pictures of periodic structures on the virtual silicon wafer from corresponding gray values of the pictures of periodic structures in the silicon wafer according to the arrangement orders, and statistical data of all the comparison difference value images; The comparison results comprise judging results of whether each comparison difference value image satisfies a preset normal distribution.

17. A method for generating a virtual silicon die, the method comprising: The virtual silicon wafer generating system and the virtual silicon wafer generating method according to any one of claims 1-16 are used, and the virtual silicon wafer generating method comprises: A detection device detects a plurality of silicon wafer information and stores the silicon wafer information in a detection database; A data collection unit collects characteristic information and process information of a silicon wafer from the detection database; A silicon wafer generating unit generates a virtual silicon wafer according to the characteristic information and the process information.

18. The method of claim 17, wherein: The data collection unit also collects hardware device information in the detection database.

19. The method of claim 17, wherein: After the virtual silicon wafer is generated, the virtual silicon wafer generating method further comprises: The silicon wafer judging unit judges whether the virtual silicon wafer meets the set requirement, and if yes, the generation of the virtual silicon wafer is completed; if no, the related process is modified according to the nature of the abnormal point, the virtual silicon wafer is generated again and judged until the set requirement is met.

Images